I haven't got a definitive answer, but what we have is a carbon double bond. Around these double bonds are areas of shielding and of deshielding. It may be that the larger S is changing this distribution.
Also S has many more electrons than O so as chemical shift is about electron motion opposing the applied field S having more electrons to do this is perhaps the answer.
p Orbital overlap between C and S is less efficient than in C and O because of a size mismatch. Thus, the resonance contributor where there is a positive charge on C is more prevalent in C=S than in C=O. A quick and dirty QM calculation shows that the bond order (for example) for C=O in acetone is ~2, while for C=S in thioacetone is ~1.6. This may be contributing to the paradoxical phenomenon of carbon is C=S being more deshielded than that in C=O.
Yes, In C=S there is a 2p-3p interaction between C and S, and in C=O, there is a 2p-2p interaction, which is perfect. The size of the S 3p orbital is bigger than the carbon 2p orbital that's why, no perfect overlapping. so the positive charge comes on carbon. and C=S will go downfield.
I'm not sure that the premise of the question is correct. The carbon shift quoted is for amides/carboxylates. Ketone/aldehyde carbonyl chemical shifts are closer to 200-220ppm.